| blob | 2c152359cdceb9b68959e9fa36b6d2c8402fad6e |
1 <?php
2 /**
3 * PHP parser that converts wiki markup to HTML.
4 *
5 * This program is free software; you can redistribute it and/or modify
6 * it under the terms of the GNU General Public License as published by
7 * the Free Software Foundation; either version 2 of the License, or
8 * (at your option) any later version.
9 *
10 * This program is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License for more details.
14 *
15 * You should have received a copy of the GNU General Public License along
16 * with this program; if not, write to the Free Software Foundation, Inc.,
17 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
18 * http://www.gnu.org/copyleft/gpl.html
19 *
20 * @file
21 * @ingroup Parser
22 */
27 /**
28 * @defgroup Parser Parser
29 */
31 /**
32 * PHP Parser - Processes wiki markup (which uses a more user-friendly
33 * syntax, such as "[[link]]" for making links), and provides a one-way
34 * transformation of that wiki markup it into (X)HTML output / markup
35 * (which in turn the browser understands, and can display).
36 *
37 * There are seven main entry points into the Parser class:
38 *
39 * - Parser::parse()
40 * produces HTML output
41 * - Parser::preSaveTransform()
42 * produces altered wiki markup
43 * - Parser::preprocess()
44 * removes HTML comments and expands templates
45 * - Parser::cleanSig() and Parser::cleanSigInSig()
46 * cleans a signature before saving it to preferences
47 * - Parser::getSection()
48 * return the content of a section from an article for section editing
49 * - Parser::replaceSection()
50 * replaces a section by number inside an article
51 * - Parser::getPreloadText()
52 * removes <noinclude> sections and <includeonly> tags
53 *
54 * Globals used:
55 * object: $wgContLang
56 *
57 * @warning $wgUser or $wgTitle or $wgRequest or $wgLang. Keep them away!
58 *
59 * @par Settings:
60 * $wgNamespacesWithSubpages
61 *
62 * @par Settings only within ParserOptions:
63 * $wgAllowExternalImages
64 * $wgAllowSpecialInclusion
65 * $wgInterwikiMagic
66 * $wgMaxArticleSize
67 *
68 * @ingroup Parser
69 */
71 /**
72 * Update this version number when the ParserOutput format
73 * changes in an incompatible way, so the parser cache
74 * can automatically discard old data.
75 */
78 /**
79 * Update this version number when the output of serialiseHalfParsedText()
80 * changes in an incompatible way
81 */
84 # Flags for Parser::setFunctionHook
88 # Constants needed for external link processing
89 # Everything except bracket, space, or control characters
90 # \p{Zs} is unicode 'separator, space' category. It covers the space 0x20
91 # as well as U+3000 is IDEOGRAPHIC SPACE for bug 19052
93 # Simplified expression to match an IPv4 or IPv6 address, or
94 # at least one character of a host name (embeds EXT_LINK_URL_CLASS)
96 # RegExp to make image URLs (embeds IPv6 part of EXT_LINK_ADDR)
97 // @codingStandardsIgnoreStart Generic.Files.LineLength
98 const EXT_IMAGE_REGEX = '/^(http:\/\/|https:\/\/)((?:\\[(?i:[0-9a-f:.]+)\\])?[^][<>"\\x00-\\x20\\x7F\p{Zs}]+)
100 // @codingStandardsIgnoreEnd
102 # Regular expression for a non-newline space
105 # Flags for preprocessToDom
108 # Allowed values for $this->mOutputType
109 # Parameter to startExternalParse().
116 /**
117 * @var string Prefix and suffix for temporary replacement strings
118 * for the multipass parser.
119 *
120 * \x7f should never appear in input as it's disallowed in XML.
121 * Using it at the front also gives us a little extra robustness
122 * since it shouldn't match when butted up against identifier-like
123 * string constructs.
124 *
125 * Must not consist of all title characters, or else it will change
126 * the behavior of <nowiki> in a link.
127 *
128 * Must have a character that needs escaping in attributes, otherwise
129 * someone could put a strip marker in an attribute, to get around
130 * escaping quote marks, and break out of the attribute. Thus we add
131 * `'".
132 */
136 # Markers used for wrapping the table of contents
140 # Persistent:
154 # Initialised by initialiseVariables()
156 /**
157 * @var MagicWordArray
158 */
161 /**
162 * @var MagicWordArray
163 */
165 # Initialised in constructor
168 # Initialized in getPreprocessor()
169 /** @var Preprocessor */
172 # Cleared with clearState():
173 /**
174 * @var ParserOutput
175 */
179 /**
180 * @var StripState
181 */
185 /**
186 * @var LinkHolderArray
187 */
197 /**
198 * @var User
199 */
202 # Temporary
203 # These are variables reset at least once per parse regardless of $clearState
205 /**
206 * @var ParserOptions
207 */
210 /**
211 * @var Title
212 */
224 /**
225 * @var string Deprecated accessor for the strip marker prefix.
226 * @deprecated since 1.26; use Parser::MARKER_PREFIX instead.
227 **/
230 /**
231 * @var array Array with the language name of each language link (i.e. the
232 * interwiki prefix) in the key, value arbitrary. Used to avoid sending
233 * duplicate language links to the ParserOutput.
234 */
237 /**
238 * @var MapCacheLRU|null
239 * @since 1.24
240 *
241 * A cache of the current revisions of titles. Keys are $title->getPrefixedDbKey()
242 */
245 /**
246 * @var bool Recursive call protection.
247 * This variable should be treated as if it were private.
248 */
251 /** @var SectionProfiler */
254 /**
255 * @var LinkRenderer
256 */
259 /**
260 * @param array $conf
261 */
271 # Preprocessor_Hash is much faster than Preprocessor_DOM under HipHop
274 # PECL extension that conflicts with the core DOM extension (bug 13770)
281 }
283 }
285 /**
286 * Reduce memory usage to reduce the impact of circular references
287 */
291 }
294 }
295 }
297 /**
298 * Allow extensions to clean up when the parser is cloned
299 */
303 // Bug 56226: When you create a reference "to" an object field, that
304 // makes the object field itself be a reference too (until the other
305 // reference goes out of scope). When cloning, any field that's a
306 // reference is copied as a reference in the new object. Both of these
307 // are defined PHP5 behaviors, as inconvenient as it is for us when old
308 // hooks from PHP4 days are passing fields by reference.
310 // Make a non-reference copy of the field, then rebind the field to
311 // reference the new copy.
315 }
318 }
320 /**
321 * Do various kinds of initialisation on the first call of the parser
322 */
326 }
334 }
336 /**
337 * Clear Parser state
338 *
339 * @private
340 */
344 }
360 # Clear these on every parse, bug 4549
368 ];
377 # Fix cloning
380 }
385 }
387 /**
388 * Convert wikitext to HTML
389 * Do not call this function recursively.
390 *
391 * @param string $text Text we want to parse
392 * @param Title $title
393 * @param ParserOptions $options
394 * @param bool $linestart
395 * @param bool $clearState
396 * @param int $revid Number to pass in {{REVISIONID}}
397 * @return ParserOutput A ParserOutput
398 */
402 ) {
403 /**
404 * First pass--just handle <nowiki> sections, pass the rest off
405 * to internalParse() which does all the real work.
406 */
411 // We use U+007F DELETE to construct strip markers, so we have to make
412 // sure that this character does not occur in the input text.
415 }
423 }
436 }
439 # No more strip!
446 /**
447 * A converted title will be provided in the output object if title and
448 * content conversion are enabled, the article text does not contain
449 * a conversion-suppressing double-underscore tag, and no
450 * {{DISPLAYTITLE:...}} is present. DISPLAYTITLE takes precedence over
451 * automatic link conversion.
452 */
457 ) {
464 }
465 }
467 # Done parsing! Compute runtime adaptive expiry if set
470 # Warn if too many heavyweight parser functions were used
475 );
476 }
478 # Information on include size limits, for the benefit of users who try to skirt them
486 );
487 }
492 );
496 );
499 );
502 );
505 );
508 );
512 );
518 // Sanitize for comment. Note '‐' in the replacement is U+2010,
519 // which looks much like the problematic '-'.
522 }
524 // Add on template profiling data in human/machine readable way
528 } );
533 }
536 // Add other cache related metadata
539 }
549 ) {
552 }
553 }
565 }
567 /**
568 * Half-parse wikitext to half-parsed HTML. This recursive parser entry point
569 * can be called from an extension tag hook.
570 *
571 * The output of this function IS NOT SAFE PARSED HTML; it is "half-parsed"
572 * instead, which means that lists and links have not been fully parsed yet,
573 * and strip markers are still present.
574 *
575 * Use recursiveTagParseFully() to fully parse wikitext to output-safe HTML.
576 *
577 * Use this function if you're a parser tag hook and you want to parse
578 * wikitext before or after applying additional transformations, and you
579 * intend to *return the result as hook output*, which will cause it to go
580 * through the rest of parsing process automatically.
581 *
582 * If $frame is not provided, then template variables (e.g., {{{1}}}) within
583 * $text are not expanded
584 *
585 * @param string $text Text extension wants to have parsed
586 * @param bool|PPFrame $frame The frame to use for expanding any template variables
587 * @return string UNSAFE half-parsed HTML
588 */
594 }
596 /**
597 * Fully parse wikitext to fully parsed HTML. This recursive parser entry
598 * point can be called from an extension tag hook.
599 *
600 * The output of this function is fully-parsed HTML that is safe for output.
601 * If you're a parser tag hook, you might want to use recursiveTagParse()
602 * instead.
603 *
604 * If $frame is not provided, then template variables (e.g., {{{1}}}) within
605 * $text are not expanded
606 *
607 * @since 1.25
608 *
609 * @param string $text Text extension wants to have parsed
610 * @param bool|PPFrame $frame The frame to use for expanding any template variables
611 * @return string Fully parsed HTML
612 */
617 }
619 /**
620 * Expand templates and variables in the text, producing valid, static wikitext.
621 * Also removes comments.
622 * Do not call this function recursively.
623 * @param string $text
624 * @param Title $title
625 * @param ParserOptions $options
626 * @param int|null $revid
627 * @param bool|PPFrame $frame
628 * @return mixed|string
629 */
632 ) {
637 }
643 }
645 /**
646 * Recursive parser entry point that can be called from an extension tag
647 * hook.
648 *
649 * @param string $text Text to be expanded
650 * @param bool|PPFrame $frame The frame to use for expanding any template variables
651 * @return string
652 * @since 1.19
653 */
658 }
660 /**
661 * Process the wikitext for the "?preload=" feature. (bug 5210)
662 *
663 * "<noinclude>", "<includeonly>" etc. are parsed as for template
664 * transclusion, comments, templates, arguments, tags hooks and parser
665 * functions are untouched.
666 *
667 * @param string $text
668 * @param Title $title
669 * @param ParserOptions $options
670 * @param array $params
671 * @return string
672 */
677 # Parser (re)initialisation
686 }
688 /**
689 * Get a random string
690 *
691 * @return string
692 * @deprecated since 1.26; use wfRandomString() instead.
693 */
697 }
699 /**
700 * Set the current user.
701 * Should only be used when doing pre-save transform.
702 *
703 * @param User|null $user User object or null (to reset)
704 */
707 }
709 /**
710 * Accessor for mUniqPrefix.
711 *
712 * @return string
713 * @deprecated since 1.26; use Parser::MARKER_PREFIX instead.
714 */
718 }
720 /**
721 * Set the context title
722 *
723 * @param Title $t
724 */
728 }
731 # Strip the fragment to avoid various odd effects
735 }
736 }
738 /**
739 * Accessor for the Title object
740 *
741 * @return Title
742 */
745 }
747 /**
748 * Accessor/mutator for the Title object
749 *
750 * @param Title $x Title object or null to just get the current one
751 * @return Title
752 */
755 }
757 /**
758 * Set the output type
759 *
760 * @param int $ot New value
761 */
764 # Shortcut alias
770 ];
771 }
773 /**
774 * Accessor/mutator for the output type
775 *
776 * @param int|null $x New value or null to just get the current one
777 * @return int
778 */
781 }
783 /**
784 * Get the ParserOutput object
785 *
786 * @return ParserOutput
787 */
790 }
792 /**
793 * Get the ParserOptions object
794 *
795 * @return ParserOptions
796 */
799 }
801 /**
802 * Accessor/mutator for the ParserOptions object
803 *
804 * @param ParserOptions $x New value or null to just get the current one
805 * @return ParserOptions Current ParserOptions object
806 */
809 }
811 /**
812 * @return int
813 */
816 }
818 /**
819 * @param int $id
820 */
823 }
825 /**
826 * Get a language object for use in parser functions such as {{FORMATNUM:}}
827 * @return Language
828 */
831 }
833 /**
834 * Get the target language for the content being parsed. This is usually the
835 * language that the content is in.
836 *
837 * @since 1.19
838 *
839 * @throws MWException
840 * @return Language
841 */
851 }
854 }
856 /**
857 * Get the language object for language conversion
858 * @return Language|null
859 */
862 }
864 /**
865 * Get a User object either from $this->mUser, if set, or from the
866 * ParserOptions object otherwise
867 *
868 * @return User
869 */
873 }
875 }
877 /**
878 * Get a preprocessor object
879 *
880 * @return Preprocessor
881 */
886 }
888 }
890 /**
891 * Get a LinkRenderer instance to make links with
892 *
893 * @since 1.28
894 * @return LinkRenderer
895 */
902 );
903 }
906 }
908 /**
909 * Replaces all occurrences of HTML-style comments and the given tags
910 * in the text with a random marker and returns the next text. The output
911 * parameter $matches will be an associative array filled with data in
912 * the form:
913 *
914 * @code
915 * 'UNIQ-xxxxx' => [
916 * 'element',
917 * 'tag content',
918 * [ 'param' => 'x' ],
919 * '<element param="x">tag content</element>' ]
920 * @endcode
921 *
922 * @param array $elements List of element names. Comments are always extracted.
923 * @param string $text Source text string.
924 * @param array $matches Out parameter, Array: extracted tags
925 * @param string|null $uniq_prefix
926 * @return string Stripped text
927 * @since 1.26 The uniq_prefix argument is deprecated.
928 */
929 public static function extractTagsAndParams( $elements, $text, &$matches, $uniq_prefix = null ) {
932 }
945 }
947 # comment
953 # tag
958 }
964 # Empty element tag, <tag />
973 }
977 # No end tag -- let it run out to the end of the text.
983 }
984 }
990 }
992 }
994 /**
995 * Get a list of strippable XML-like elements
996 *
997 * @return array
998 */
1001 }
1003 /**
1004 * Add an item to the strip state
1005 * Returns the unique tag which must be inserted into the stripped text
1006 * The tag will be replaced with the original text in unstrip()
1007 *
1008 * @param string $text
1009 *
1010 * @return string
1011 */
1017 }
1019 /**
1020 * parse the wiki syntax used to render tables
1021 *
1022 * @private
1023 * @param string $text
1024 * @return string
1025 */
1043 }
1050 # First check if we are starting a new table
1063 # Don't do any of the following
1067 # We are ending a table
1073 }
1077 }
1081 }
1085 # Now we have a table row
1088 # Whats after the tag is now only attributes
1101 }
1105 }
1114 ) {
1115 # This might be cell elements, td, th or captions
1121 }
1123 // Implies both are valid for table headings.
1126 }
1128 # Split up multiple cells on the same line.
1129 # FIXME : This can result in improper nesting of tags processed
1130 # by earlier parser steps.
1135 # Loop through each table cell
1142 }
1147 }
1153 }
1163 }
1167 # A cell could contain both parameters and data
1170 # Bug 553: Note that a '|' inside an invalid link should not
1171 # be mistaken as delimiting cell parameters
1180 }
1184 }
1185 }
1187 }
1189 # Closing open td, tr && table
1193 }
1196 }
1199 }
1202 }
1204 # Remove trailing line-ending (b/c)
1207 }
1209 # special case: don't return empty table
1212 }
1215 }
1217 /**
1218 * Helper function for parse() that transforms wiki markup into half-parsed
1219 * HTML. Only called for $mOutputType == self::OT_HTML.
1220 *
1221 * @private
1222 *
1223 * @param string $text The text to parse
1224 * @param bool $isMain Whether this is being called from the main parse() function
1225 * @param PPFrame|bool $frame A pre-processor frame
1226 *
1227 * @return string
1228 */
1233 # Hook to suspend the parser in this state
1236 }
1238 # if $frame is provided, then use $frame for replacing any variables
1240 # use frame depth to infer how include/noinclude tags should be handled
1241 # depth=0 means this is the top-level document; otherwise it's an included document
1246 }
1250 # if $frame is not provided, then use old-style replaceVariables
1252 }
1260 [],
1262 );
1265 # Tables need to come after variable replacement for things to work
1266 # properly; putting them before other transformations should keep
1267 # exciting things like link expansions from showing up in surprising
1268 # places.
1280 # replaceInternalLinks may sometimes leave behind
1281 # absolute URLs, which have to be masked to hide them from replaceExternalLinks
1288 }
1290 /**
1291 * Helper function for parse() that transforms half-parsed HTML into fully
1292 * parsed HTML.
1293 *
1294 * @param string $text
1295 * @param bool $isMain
1296 * @param bool $linestart
1297 * @return string
1298 */
1304 }
1306 # Clean up special characters, only run once, next-to-last before doBlockLevels
1308 # french spaces, last one Guillemet-left
1309 # only if there is something before the space
1311 # french spaces, Guillemet-right
1314 ];
1321 /**
1322 * The input doesn't get language converted if
1323 * a) It's disabled
1324 * b) Content isn't converted
1325 * c) It's a conversion table
1326 * d) it is an interface message (which is in the user language)
1327 */
1330 ) {
1332 # The position of the convert() call should not be changed. it
1333 # assumes that the links are all replaced and the only thing left
1334 # is the <nowiki> mark.
1336 }
1337 }
1343 }
1353 }
1355 # attempt to sanitize at least some nesting problems
1356 # (bug #2702 and quite a few others)
1358 # ''Something [http://www.cool.com cool''] -->
1359 # <i>Something</i><a href="http://www.cool.com"..><i>cool></i></a>
1362 # fix up an anchor inside another anchor, only
1363 # at least for a single single nested link (bug 3695)
1366 # fix div inside inline elements- doBlockLevels won't wrap a line which
1367 # contains a div, so fix it up here; replace
1368 # div with escaped text
1371 # remove empty italic or bold tag pairs, some
1372 # introduced by rules above
1374 ];
1380 }
1384 }
1387 }
1389 /**
1390 * Replace special strings like "ISBN xxx" and "RFC xxx" with
1391 * magic external links.
1392 *
1393 * DML
1394 * @private
1395 *
1396 * @param string $text
1397 *
1398 * @return string
1399 */
1408 '!(?: # Start cases
1410 (<.*?>) | # m[2]: Skip stuff inside
1413 # m[4]: Post-protocol path
1419 [0-9Xx] # check digit
1423 }
1425 /**
1426 * @throws MWException
1427 * @param array $m
1428 * @return HTML|string
1429 */
1432 # Skip anchor
1435 # Skip HTML element
1438 # Free external link
1441 # RFC or PMID
1445 }
1454 }
1463 }
1466 return Linker::makeExternalLink( $url, "{$keyword} {$id}", true, $cssClass, [], $this->mTitle );
1469 ) {
1470 # ISBN
1478 ] );
1483 [
1486 ]
1487 );
1490 }
1491 }
1493 /**
1494 * Make a free external link, given a user-supplied URL
1495 *
1496 * @param string $url
1497 * @param int $numPostProto
1498 * The number of characters after the protocol.
1499 * @return string HTML
1500 * @private
1501 */
1505 # The characters '<' and '>' (which were escaped by
1506 # removeHTMLtags()) should not be included in
1507 # URLs, per RFC 2396.
1508 # Make terminate a URL as well (bug T84937)
1514 PREG_OFFSET_CAPTURE
1515 ) ) {
1518 }
1520 # Move trailing punctuation to $trail
1522 # If there is no left bracket, then consider right brackets fair game too
1525 }
1529 # Don't break a trailing HTML entity by moving the ; into $trail
1530 # This is in hot code, so use substr_compare to avoid having to
1531 # create a new string object for the comparison
1533 # more optimization: instead of running preg_match with a $
1534 # anchor, which can be slow, do the match on the reversed
1535 # string starting at the desired offset.
1536 # un-reversed regexp is: /&([a-z]+|#x[\da-f]+|#\d+)$/i
1539 }
1540 }
1544 }
1546 # Verify that we still have a real URL after trail removal, and
1547 # not just lone protocol
1550 }
1554 # Is this an external image?
1557 # Not an image, make a link
1562 # Register it in the output object...
1563 # Replace unnecessary URL escape codes with their equivalent characters
1566 }
1568 }
1570 /**
1571 * Parse headers and return html
1572 *
1573 * @private
1574 *
1575 * @param string $text
1576 *
1577 * @return string
1578 */
1583 }
1585 }
1587 /**
1588 * Replace single quotes with HTML markup
1589 * @private
1590 *
1591 * @param string $text
1592 *
1593 * @return string The altered text
1594 */
1600 }
1603 }
1605 /**
1606 * Helper function for doAllQuotes()
1607 *
1608 * @param string $text
1609 *
1610 * @return string
1611 */
1617 }
1619 // First, do some preliminary work. This may shift some apostrophes from
1620 // being mark-up to being text. It also counts the number of occurrences
1621 // of bold and italics mark-ups.
1626 // If there are ever four apostrophes, assume the first is supposed to
1627 // be text, and the remaining three constitute mark-up for bold text.
1628 // (bug 13227: ''''foo'''' turns into ' ''' foo ' ''')
1634 // If there are more than 5 apostrophes in a row, assume they're all
1635 // text except for the last 5.
1636 // (bug 13227: ''''''foo'''''' turns into ' ''''' foo ' ''''')
1640 }
1641 // Count the number of occurrences of bold and italics mark-ups.
1649 }
1650 }
1652 // If there is an odd number of both bold and italics, it is likely
1653 // that one of the bold ones was meant to be an apostrophe followed
1654 // by italics. Which one we cannot know for certain, but it is more
1655 // likely to be one that has a single-letter word before it.
1667 }
1670 // if $firstsingleletterword is set, we don't
1671 // look at the other options, so we can bail early.
1676 }
1677 }
1678 }
1679 }
1681 // If there is a single-letter word, use it!
1686 // If not, but there's a multi-letter word, use that one.
1690 // ... otherwise use the first one that has neither.
1691 // (notice that it is possible for all three to be -1 if, for example,
1692 // there is only one pentuple-apostrophe in the line)
1695 }
1696 }
1698 // Now let's actually convert our apostrophic mush to HTML!
1709 }
1728 }
1745 }
1765 }
1766 }
1767 }
1769 }
1770 // Now close all remaining tags. Notice that the order is important.
1773 }
1776 }
1779 }
1780 // There might be lonely ''''', so make sure we have a buffer
1783 }
1785 }
1787 /**
1788 * Replace external links (REL)
1789 *
1790 * Note: this is all very hackish and the order of execution matters a lot.
1791 * Make sure to run tests/parser/parserTests.php if you change this code.
1792 *
1793 * @private
1794 *
1795 * @param string $text
1796 *
1797 * @throws MWException
1798 * @return string
1799 */
1806 }
1816 # The characters '<' and '>' (which were escaped by
1817 # removeHTMLtags()) should not be included in
1818 # URLs, per RFC 2396.
1823 }
1825 # If the link text is an image URL, replace it with an <img> tag
1826 # This happened by accident in the original parser, but some people used it extensively
1830 }
1834 # Set linktype for CSS - if URL==text, link is essentially free
1837 # No link text, e.g. [http://domain.tld/some.link]
1839 # Autonumber
1844 # Have link text, e.g. [http://domain.tld/some.link text]s
1845 # Check for trail
1847 }
1853 # Use the encoded URL
1854 # This means that users can paste URLs directly into the text
1855 # Funny characters like ö aren't valid in URLs anyway
1856 # This was changed in August 2004
1860 # Register link in the output object.
1861 # Replace unnecessary URL escape codes with the referenced character
1862 # This prevents spammers from hiding links from the filters
1865 }
1868 }
1870 /**
1871 * Get the rel attribute for a particular external link.
1872 *
1873 * @since 1.21
1874 * @param string|bool $url Optional URL, to extract the domain from for rel =>
1875 * nofollow if appropriate
1876 * @param Title $title Optional Title, for wgNoFollowNsExceptions lookups
1877 * @return string|null Rel attribute for $url
1878 */
1884 ) {
1886 }
1888 }
1890 /**
1891 * Get an associative array of additional HTML attributes appropriate for a
1892 * particular external link. This currently may include rel => nofollow
1893 * (depending on configuration, namespace, and the URL's domain) and/or a
1894 * target attribute (depending on configuration).
1895 *
1896 * @param string $url URL to extract the domain from for rel =>
1897 * nofollow if appropriate
1898 * @return array Associative array of HTML attributes
1899 */
1908 // T133507. New windows can navigate parent cross-origin.
1909 // Including noreferrer due to lacking browser
1910 // support of noopener. Eventually noreferrer should be removed.
1913 }
1915 }
1916 }
1919 }
1921 /**
1922 * Replace unusual escape codes in a URL with their equivalent characters
1923 *
1924 * @deprecated since 1.24, use normalizeLinkUrl
1925 * @param string $url
1926 * @return string
1927 */
1931 }
1933 /**
1934 * Replace unusual escape codes in a URL with their equivalent characters
1935 *
1936 * This generally follows the syntax defined in RFC 3986, with special
1937 * consideration for HTTP query strings.
1938 *
1939 * @param string $url
1940 * @return string
1941 */
1943 # First, make sure unsafe characters are encoded
1947 },
1948 $url
1949 );
1954 # Fragment part - 'fragment'
1960 }
1962 # Query part - 'query' minus &=+;
1968 }
1970 # Scheme and path part - 'pchar'
1971 # (we assume no userinfo or encoded colons in the host)
1976 }
1983 # Unescape it
1986 # Leave it escaped, but use uppercase for a-f
1988 }
1989 };
1991 }
1993 /**
1994 * make an image if it's allowed, either through the global
1995 * option, through the exception, or through the on-wiki whitelist
1996 *
1997 * @param string $url
1998 *
1999 * @return string
2000 */
2005 # $imagesfrom could be either a single string or an array of strings, parse out the latter
2012 }
2013 }
2018 }
2022 ) {
2024 # Image found
2026 }
2027 }
2030 ) {
2034 );
2037 # Sanitize the regex fragment, make it case-insensitive, ignore blank entries/comments
2040 }
2042 # Image matches a whitelist entry
2045 }
2046 }
2047 }
2049 }
2051 /**
2052 * Process [[ ]] wikilinks
2053 *
2054 * @param string $s
2055 *
2056 * @return string Processed text
2057 *
2058 * @private
2059 */
2063 }
2065 /**
2066 * Process [[ ]] wikilinks (RIL)
2067 * @param string $s
2068 * @throws MWException
2069 * @return LinkHolderArray
2070 *
2071 * @private
2072 */
2077 # the % is needed to support urlencoded titles as well
2080 # Match a link having the form [[namespace:link|alternate]]trail
2082 # Match cases where there is no "]]", which might still be images
2084 }
2088 # split the entire text string on occurrences of [[
2090 # get the first element (all text up to first [[), and remove the space we added
2099 # Match the end of a line for a word that's not followed by whitespace,
2100 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
2104 }
2108 }
2117 }
2120 }
2124 // @codingStandardsIgnoreStart Squiz.WhiteSpace.SemicolonSpacing.Incorrect
2125 # Loop for each link
2127 // @codingStandardsIgnoreEnd
2129 # Check for excessive memory usage
2131 # Too big
2132 # Do the existence check, replace the link holders and clear the array
2135 }
2143 }
2144 # first link
2148 }
2149 }
2155 # If we get a ] at the beginning of $m[3] that means we have a link that's something like:
2156 # [[Image:Foo.jpg|[http://example.com desc]]] <- having three ] in a row fucks up,
2157 # the real problem is with the $e1 regex
2158 # See bug 1300.
2159 # Still some problems for cases where the ] is meant to be outside punctuation,
2160 # and no image is in sight. See bug 2095.
2164 ) {
2167 }
2168 # fix up urlencoded title texts
2170 # Should anchors '#' also be rejected?
2172 }
2175 # Invalid, but might be an image with a link in its caption
2180 }
2185 }
2189 # Don't allow internal links to pages containing
2190 # PROTO: where PROTO is a valid URL protocol; these
2191 # should be external links.
2195 }
2197 # Make subpage if necessary
2202 }
2206 # Strip off leading ':'
2208 }
2215 }
2224 # look at the next 'line' to see if we can close it there
2229 }
2232 # the first ]] closes the inner link, the second the image
2238 # if there's exactly one ]] that's fine, we'll keep looking
2241 # if $next_line is invalid too, we need look no further
2244 }
2245 }
2247 # we couldn't find the end of this imageLink, so output it raw
2248 # but don't ignore what might be perfectly normal links in the text we've examined
2251 # note: no $trail, because without an end, there *is* no trail
2253 }
2256 # note: no $trail, because without an end, there *is* no trail
2258 }
2259 }
2265 # Bug 4598 madness. Handle the quotes only if they come from the alternate part
2266 # [[Lista d''e paise d''o munno]] -> <a href="...">Lista d''e paise d''o munno</a>
2267 # [[Criticism of Harry Potter|Criticism of ''Harry Potter'']]
2268 # -> <a href="Criticism of Harry Potter">Criticism of <i>Harry Potter</i></a>
2270 }
2272 # Link not escaped by : , create the various objects
2274 # Interwikis
2279 )
2280 ) {
2281 # Bug 24502: filter duplicates
2285 }
2290 }
2295 # if no parameters were passed, $text
2296 # becomes something like "File:Foo.png",
2297 # which we don't want to pass on to the
2298 # image generator
2301 # recursively parse links inside the image caption
2302 # actually, this will parse them in any other parameters, too,
2303 # but it might be hard to fix that, and it doesn't matter ATM
2306 }
2307 # cloak any absolute URLs inside the image markup, so replaceExternalLinks() won't touch them
2311 }
2319 }
2325 /**
2326 * Strip the whitespace Category links produce, see bug 87
2327 */
2331 }
2332 }
2334 # Self-link checking. For some languages, variants of the title are checked in
2335 # LinkHolderArray::doVariants() to allow batching the existence checks necessary
2336 # for linking to a different variant.
2340 }
2342 # NS_MEDIA is a pseudo-namespace for linking directly to a file
2343 # @todo FIXME: Should do batch file existence checks, see comment below
2345 # Give extensions a chance to select the file revision for us
2350 # Fetch and register the file (file title may be different via hooks)
2352 # Cloak with NOPARSE to avoid replacement in replaceExternalLinks
2356 }
2358 # Some titles, such as valid special pages or files in foreign repos, should
2359 # be shown as bluelinks even though they're not included in the page table
2360 # @todo FIXME: isAlwaysKnown() can be expensive for file links; we should really do
2361 # batch file existence checks for NS_FILE and NS_MEDIA
2366 # Links will be added to the output link list after checking
2368 }
2369 }
2371 }
2373 /**
2374 * Render a forced-blue link inline; protect against double expansion of
2375 * URLs if we're in a mode that prepends full URL prefixes to internal links.
2376 * Since this little disaster has to split off the trail text to avoid
2377 * breaking URLs in the following text without breaking trails on the
2378 * wiki links, it's been made into a horrible function.
2379 *
2380 * @param Title $nt
2381 * @param string $text
2382 * @param string $trail
2383 * @param string $prefix
2384 * @return string HTML-wikitext mix oh yuck
2385 */
2391 }
2395 );
2398 }
2400 /**
2401 * Insert a NOPARSE hacky thing into any inline links in a chunk that's
2402 * going to go through further parsing steps before inline URL expansion.
2403 *
2404 * Not needed quite as much as it used to be since free links are a bit
2405 * more sensible these days. But bracketed links are still an issue.
2406 *
2407 * @param string $text More-or-less HTML
2408 * @return string Less-or-more HTML with NOPARSE bits
2409 */
2413 }
2415 /**
2416 * Return true if subpage links should be expanded on this page.
2417 * @return bool
2418 */
2420 # Some namespaces don't allow subpages
2422 }
2424 /**
2425 * Handle link to subpage if necessary
2426 *
2427 * @param string $target The source of the link
2428 * @param string &$text The link text, modified as necessary
2429 * @return string The full name of the link
2430 * @private
2431 */
2434 }
2436 /**
2437 * Make lists from lines starting with ':', '*', '#', etc. (DBL)
2438 *
2439 * @param string $text
2440 * @param bool $linestart Whether or not this is at the start of a line.
2441 * @private
2442 * @return string The lists rendered as HTML
2443 */
2446 }
2448 /**
2449 * Return value of a magic variable (like PAGENAME)
2450 *
2451 * @private
2452 *
2453 * @param int $index
2454 * @param bool|PPFrame $frame
2455 *
2456 * @throws MWException
2457 * @return string
2458 */
2464 // If no title set, bad things are going to happen
2465 // later. Title should always be set since this
2466 // should only be called in the middle of a parse
2467 // operation (but the unit-tests do funky stuff)
2470 }
2472 /**
2473 * Some of these require message or data lookups and can be
2474 * expensive to check many times.
2475 */
2479 }
2480 }
2525 $value = $pageLang->getMonthAbbreviation( MWTimestamp::getLocalInstance( $ts )->format( 'n' ) );
2559 ) ) );
2569 ) ) );
2577 }
2585 }
2598 # 0 means the page doesn't exist in the database,
2599 # which means the user is previewing a new page.
2600 # The vary-revision flag must be set, because the magic word
2601 # will have a different value once the page is saved.
2604 }
2608 # Let the edit saving system know we should parse the page
2609 # *after* a revision ID has been assigned.
2616 }
2619 # Let the edit saving system know we should parse the page
2620 # *after* a revision ID has been assigned. This is for null edits.
2626 # Let the edit saving system know we should parse the page
2627 # *after* a revision ID has been assigned. This is for null edits.
2633 # Let the edit saving system know we should parse the page
2634 # *after* a revision ID has been assigned. This is for null edits.
2640 # Let the edit saving system know we should parse the page
2641 # *after* a revision ID has been assigned. This is for null edits.
2647 # Let the edit saving system know we should parse the page
2648 # *after* a revision ID has been assigned. This is for null edits.
2654 # Let the edit saving system know we should parse the page
2655 # *after* a revision ID has been assigned. This is for null edits.
2661 # Let the edit saving system know we should parse the page
2662 # *after* a revision ID has been assigned for null edits.
2706 # @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
2707 # int to remove the padding
2716 );
2725 false
2726 );
2732 # @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
2733 # int to remove the padding
2794 );
2797 }
2801 }
2804 }
2806 /**
2807 * initialise the magic variables (like CURRENTMONTHNAME) and substitution modifiers
2808 *
2809 * @private
2810 */
2817 }
2819 /**
2820 * Preprocess some wikitext and return the document tree.
2821 * This is the ghost of replace_variables().
2822 *
2823 * @param string $text The text to parse
2824 * @param int $flags Bitwise combination of:
2825 * - self::PTD_FOR_INCLUSION: Handle "<noinclude>" and "<includeonly>" as if the text is being
2826 * included. Default is to assume a direct page view.
2827 *
2828 * The generated DOM tree must depend only on the input text and the flags.
2829 * The DOM tree must be the same in OT_HTML and OT_WIKI mode, to avoid a regression of bug 4899.
2830 *
2831 * Any flag added to the $flags parameter here, or any other parameter liable to cause a
2832 * change in the DOM tree for a given text, must be passed through the section identifier
2833 * in the section edit link and thus back to extractSections().
2834 *
2835 * The output of this function is currently only cached in process memory, but a persistent
2836 * cache may be implemented at a later date which takes further advantage of these strict
2837 * dependency requirements.
2838 *
2839 * @return PPNode
2840 */
2844 }
2846 /**
2847 * Return a three-element array: leading whitespace, string contents, trailing whitespace
2848 *
2849 * @param string $s
2850 *
2851 * @return array
2852 */
2862 }
2864 }
2866 /**
2867 * Replace magic variables, templates, and template arguments
2868 * with the appropriate text. Templates are substituted recursively,
2869 * taking care to avoid infinite loops.
2870 *
2871 * Note that the substitution depends on value of $mOutputType:
2872 * self::OT_WIKI: only {{subst:}} templates
2873 * self::OT_PREPROCESS: templates but not extension tags
2874 * self::OT_HTML: all templates and extension tags
2875 *
2876 * @param string $text The text to transform
2877 * @param bool|PPFrame $frame Object describing the arguments passed to the
2878 * template. Arguments may also be provided as an associative array, as
2879 * was the usual case before MW1.12. Providing arguments this way may be
2880 * useful for extensions wishing to perform variable replacement
2881 * explicitly.
2882 * @param bool $argsOnly Only do argument (triple-brace) expansion, not
2883 * double-brace expansion.
2884 * @return string
2885 */
2887 # Is there any text? Also, Prevent too big inclusions!
2891 }
2899 }
2906 }
2908 /**
2909 * Clean up argument array - refactored in 1.9 so parserfunctions can use it, too.
2910 *
2911 * @param array $args
2912 *
2913 * @return array
2914 */
2927 }
2930 }
2931 }
2932 }
2935 }
2937 /**
2938 * Warn the user when a parser limitation is reached
2939 * Will warn at most once the user per limitation type
2940 *
2941 * The results are shown during preview and run through the Parser (See EditPage.php)
2942 *
2943 * @param string $limitationType Should be one of:
2944 * 'expensive-parserfunction' (corresponding messages:
2945 * 'expensive-parserfunction-warning',
2946 * 'expensive-parserfunction-category')
2947 * 'post-expand-template-argument' (corresponding messages:
2948 * 'post-expand-template-argument-warning',
2949 * 'post-expand-template-argument-category')
2950 * 'post-expand-template-inclusion' (corresponding messages:
2951 * 'post-expand-template-inclusion-warning',
2952 * 'post-expand-template-inclusion-category')
2953 * 'node-count-exceeded' (corresponding messages:
2954 * 'node-count-exceeded-warning',
2955 * 'node-count-exceeded-category')
2956 * 'expansion-depth-exceeded' (corresponding messages:
2957 * 'expansion-depth-exceeded-warning',
2958 * 'expansion-depth-exceeded-category')
2959 * @param string|int|null $current Current value
2960 * @param string|int|null $max Maximum allowed, when an explicit limit has been
2961 * exceeded, provide the values (optional)
2962 */
2964 # does no harm if $current and $max are present but are unnecessary for the message
2965 # Not doing ->inLanguage( $this->mOptions->getUserLangObj() ), since this is shown
2966 # only during preview, and that would split the parser cache unnecessarily.
2971 }
2973 /**
2974 * Return the text of a template, after recursively
2975 * replacing any variables or templates within the template.
2976 *
2977 * @param array $piece The parts of the template
2978 * $piece['title']: the title, i.e. the part before the |
2979 * $piece['parts']: the parameter array
2980 * $piece['lineStart']: whether the brace was at the start of a line
2981 * @param PPFrame $frame The current frame, contains template arguments
2982 * @throws Exception
2983 * @return string The text of the template
2984 */
2987 // Flags
2989 // $text has been filled
2991 // wiki markup in $text should be escaped
2993 // $text is HTML, armour it against wikitext transformation
2995 // Force interwiki transclusion to be done in raw mode not rendered
2997 // $text is a DOM node needing expansion in a child frame
2999 // $text is a DOM node needing expansion in the current frame
3002 # Title object, where $text came from
3005 # $part1 is the bit before the first |, and must contain only title characters.
3006 # Various prefixes will be stripped from it later.
3011 # Original title text preserved for various purposes
3014 # $args is a list of argument nodes, starting from index 0, not including $part1
3015 # @todo FIXME: If piece['parts'] is null then the call to getLength()
3016 # below won't work b/c this $args isn't an object
3021 # SUBST
3025 # Possibilities for substMatch: "subst", "safesubst" or FALSE
3026 # Decide whether to expand template or keep wikitext as-is.
3032 }
3038 }
3039 }
3044 }
3045 }
3047 # Variables
3054 }
3056 }
3057 }
3059 # MSG, MSGNW and RAW
3061 # Check for MSGNW:
3066 # Remove obsolete MSG:
3069 }
3071 # Check for RAW:
3075 }
3076 }
3078 # Parser functions
3087 }
3092 }
3094 # The interface for parser functions allows for extracting
3095 # flags into the local scope. Extract any forwarded flags
3096 # here.
3098 }
3099 }
3101 # Finish mangling title and then check for loops.
3102 # Set $title to a Title object and $titleText to the PDBK
3105 # Split the title into page and subpage
3111 }
3115 # Check for language variants if the template is not found
3118 }
3119 # Do recursion depth check
3127 }
3128 }
3129 }
3131 # Load from database
3138 ) {
3140 // Pass the template arguments as URL parameters.
3141 // "uselang" will have no effect since the Language object
3142 // is forced to the one defined in ParserOptions.
3151 }
3152 }
3154 // Create a new context to execute the special page
3161 // If this page is cached, then we better not be per user.
3163 }
3175 );
3176 }
3177 }
3187 }
3188 }
3190 # If the title is valid but undisplayable, make a link to it
3194 }
3196 # Interwiki transclusion
3202 # Preprocess it like a template
3205 }
3207 }
3209 # Do infinite loop check
3210 # This has to be done after redirect resolution to avoid infinite loops via redirects
3217 }
3218 }
3220 # If we haven't found text to substitute by now, we're done
3221 # Recover the source wikitext and return it
3226 }
3228 }
3230 # Expand DOM-style return values in a child frame
3232 # Clean up argument array
3238 # Expansion is eligible for the empty-frame cache
3241 # Uncached expansion
3243 }
3244 }
3248 }
3252 }
3254 # Replace raw HTML by a placeholder
3258 # Escape nowiki-style return values
3263 ) {
3264 # Bug 529: if the template begins with a table or block-level
3265 # element, it should be treated as beginning a new line.
3266 # This behavior is somewhat controversial.
3268 }
3271 # Error, oversize inclusion
3273 # Make a working, properly escaped link if possible (bug 23588)
3276 # This will probably not be a working link, but at least it may
3277 # provide some hint of where the problem is
3280 }
3284 }
3290 }
3293 }
3295 /**
3296 * Call a parser function and return an array with text and flags.
3297 *
3298 * The returned array will always contain a boolean 'found', indicating
3299 * whether the parser function was found or not. It may also contain the
3300 * following:
3301 * text: string|object, resulting wikitext or PP DOM object
3302 * isHTML: bool, $text is HTML, armour it against wikitext transformation
3303 * isChildObj: bool, $text is a DOM node needing expansion in a child frame
3304 * isLocalObj: bool, $text is a DOM node needing expansion in the current frame
3305 * nowiki: bool, wiki markup in $text should be escaped
3306 *
3307 * @since 1.21
3308 * @param PPFrame $frame The current frame, contains template arguments
3309 * @param string $function Function name
3310 * @param array $args Arguments to the function
3311 * @throws MWException
3312 * @return array
3313 */
3317 # Case sensitive functions
3321 # Case insensitive functions
3327 }
3328 }
3332 # Workaround for PHP bug 35229 and similar
3335 }
3339 # Convert arguments to PPNodes and collect for appending to $allArgs
3346 }
3347 }
3349 # Add a frame parameter, and pass the arguments as an array
3353 # Convert arguments to plain text and append to $allArgs
3361 }
3362 }
3363 }
3367 # The interface for function hooks allows them to return a wikitext
3368 # string or an array containing the string and any flags. This mungs
3369 # things around to match what this method should return.
3374 ];
3378 }
3382 ];
3383 }
3389 }
3392 }
3397 }
3400 }
3402 /**
3403 * Get the semi-parsed DOM representation of a template with a given title,
3404 * and its redirect destination title. Cached.
3405 *
3406 * @param Title $title
3407 *
3408 * @return array
3409 */
3418 }
3421 }
3423 # Cache miss, go to the database
3429 }
3437 }
3440 }
3442 /**
3443 * Fetch the current revision of a given title. Note that the revision
3444 * (and even the title) may not exist in the database, so everything
3445 * contributing to the output of the parser should use this method
3446 * where possible, rather than getting the revisions themselves. This
3447 * method also caches its results, so using it benefits performance.
3448 *
3449 * @since 1.24
3450 * @param Title $title
3451 * @return Revision
3452 */
3457 }
3460 // Defaults to Parser::statelessFetchRevision()
3462 );
3463 }
3465 }
3467 /**
3468 * Wrapper around Revision::newFromTitle to allow passing additional parameters
3469 * without passing them on to it.
3470 *
3471 * @since 1.24
3472 * @param Title $title
3473 * @param Parser|bool $parser
3474 * @return Revision|bool False if missing
3475 */
3483 }
3486 }
3488 /**
3489 * Fetch the unparsed text of a template and register a reference to it.
3490 * @param Title $title
3491 * @return array ( string or false, Title )
3492 */
3494 // Defaults to Parser::statelessFetchTemplate()
3497 // We use U+007F DELETE to distinguish strip markers from regular text.
3501 }
3507 // If we transclude ourselves, the final result
3508 // will change based on the new version of the page
3510 }
3511 }
3512 }
3514 }
3516 /**
3517 * Fetch the unparsed text of a template and register a reference to it.
3518 * @param Title $title
3519 * @return string|bool
3520 */
3523 }
3525 /**
3526 * Static function to get a template
3527 * Can be overridden via ParserOptions::setTemplateCallback().
3528 *
3529 * @param Title $title
3530 * @param bool|Parser $parser
3531 *
3532 * @return array
3533 */
3539 # Loop to fetch the article, with up to 1 redirect
3540 // @codingStandardsIgnoreStart Generic.CodeAnalysis.ForLoopWithTestFunctionCall.NotAllowed
3542 // @codingStandardsIgnoreEnd
3543 # Give extensions a chance to select the revision instead
3554 ];
3556 }
3557 # Get the revision
3564 }
3566 # If there is no current revision, there is no page
3570 }
3577 # We fetched a rev from a different title; register it too...
3582 }
3591 }
3598 }
3603 }
3606 }
3607 # Redirect?
3610 }
3615 }
3617 /**
3618 * Fetch a file and its title and register a reference to it.
3619 * If 'broken' is a key in $options then the file will appear as a broken thumbnail.
3620 * @param Title $title
3621 * @param array $options Array of options to RepoGroup::findFile
3622 * @return File|bool
3623 */
3626 }
3628 /**
3629 * Fetch a file and its title and register a reference to it.
3630 * If 'broken' is a key in $options then the file will appear as a broken thumbnail.
3631 * @param Title $title
3632 * @param array $options Array of options to RepoGroup::findFile
3633 * @return array ( File or false, Title of file )
3634 */
3640 # Register the file as a dependency...
3643 # Update fetched file title
3646 }
3648 }
3650 /**
3651 * Helper function for fetchFileAndTitle.
3652 *
3653 * Also useful if you need to fetch a file but not use it yet,
3654 * for example to get the file's handler.
3655 *
3656 * @param Title $title
3657 * @param array $options Array of options to RepoGroup::findFile
3658 * @return File|bool
3659 */
3667 }
3669 }
3671 /**
3672 * Transclude an interwiki link.
3673 *
3674 * @param Title $title
3675 * @param string $action
3676 *
3677 * @return string
3678 */
3684 }
3690 }
3692 }
3694 /**
3695 * @param string $url
3696 * @return mixed|string
3697 */
3706 }
3713 // Though we failed to fetch the content, this status is useless.
3718 }
3725 ] );
3727 }
3729 /**
3730 * Triple brace replacement -- used for template arguments
3731 * @private
3732 *
3733 * @param array $piece
3734 * @param PPFrame $frame
3735 *
3736 * @return array
3737 */
3750 )
3751 ) {
3752 # No match in frame, use the supplied default
3754 }
3758 }
3761 # No match anywhere
3763 }
3766 }
3771 }
3774 }
3776 /**
3777 * Return the text to be used for a given extension tag.
3778 * This is the ghost of strip().
3779 *
3780 * @param array $params Associative array of parameters:
3781 * name PPNode for the tag name
3782 * attr PPNode for unparsed text where tag attributes are thought to be
3783 * attributes Optional associative array of parsed attributes
3784 * inner Contents of extension element
3785 * noClose Original text did not have a close tag
3786 * @param PPFrame $frame
3787 *
3788 * @throws MWException
3789 * @return string
3790 */
3797 // Probably expansion depth or node count exceeded. Just punt the
3798 // error up.
3800 }
3804 // See above
3806 }
3810 // See above
3812 }
3823 }
3829 }
3832 # Workaround for PHP bug 35229 and similar
3835 }
3842 }
3848 }
3851 # Extract flags to local scope (to override $markerType)
3856 }
3860 }
3865 }
3866 }
3872 // See above
3874 }
3876 }
3877 }
3887 }
3889 }
3891 /**
3892 * Increment an include size counter
3893 *
3894 * @param string $type The type of expansion
3895 * @param int $size The size of the text
3896 * @return bool False if this inclusion would take it over the maximum, true otherwise
3897 */
3904 }
3905 }
3907 /**
3908 * Increment the expensive function count
3909 *
3910 * @return bool False if the limit has been exceeded
3911 */
3915 }
3917 /**
3918 * Strip double-underscore items like __NOGALLERY__ and __NOTOC__
3919 * Fills $this->mDoubleUnderscores, returns the modified text
3920 *
3921 * @param string $text
3922 *
3923 * @return string
3924 */
3927 # The position of __TOC__ needs to be recorded
3933 # Set a placeholder. At the end we'll fill it in with the TOC.
3936 # Only keep the first one.
3938 }
3940 # Now match and remove the rest of them
3946 }
3949 }
3952 ) {
3954 }
3955 # (bug 8068) Allow control over whether robots index a page.
3956 # @todo FIXME: Bug 14899: __INDEX__ always overrides __NOINDEX__ here! This
3957 # is not desirable, the last one on the page should win.
3961 }
3965 }
3967 # Cache all double underscores in the database
3970 }
3973 }
3975 /**
3976 * @see ParserOutput::addTrackingCategory()
3977 * @param string $msg Message key
3978 * @return bool Whether the addition was successful
3979 */
3982 }
3984 /**
3985 * This function accomplishes several tasks:
3986 * 1) Auto-number headings if that option is enabled
3987 * 2) Add an [edit] link to sections for users who have enabled the option and can edit the page
3988 * 3) Add a Table of contents on the top for users who have enabled the option
3989 * 4) Auto-anchor headings
3990 *
3991 * It loops through all headlines, collects the necessary data, then splits up the
3992 * string and re-inserts the newly formatted headlines.
3993 *
3994 * @param string $text
3995 * @param string $origText Original, untouched wikitext
3996 * @param bool $isMain
3997 * @return mixed|string
3998 * @private
3999 */
4003 # Inhibit editsection links if requested in the page
4009 }
4012 }
4014 # Get all headlines for numbering them and adding funky stuff like [edit]
4015 # links - this is for later, but we need the number of headlines right now
4020 $matches
4021 );
4023 # if there are fewer than 4 headlines in the article, do not show TOC
4024 # unless it's been explicitly enabled.
4028 # Allow user to stipulate that a page should have a "new section"
4029 # link added via __NEWSECTIONLINK__
4032 }
4034 # Allow user to remove the "new section"
4035 # link via __NONEWSECTIONLINK__
4038 }
4040 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
4041 # override above conditions and always show TOC above first header
4045 }
4047 # headline counter
4051 # Ugh .. the TOC should have neat indentation levels which can be
4052 # passed to the skin functions. These are determined here
4086 }
4090 }
4094 # Increase TOC level
4101 }
4103 # Decrease TOC level, find level to jump to
4107 # Found last matching level
4111 # Found first matching level below current level
4114 }
4115 }
4118 }
4121 # Unindent only if the previous toc level was shown :p
4126 }
4127 }
4129 # No change in level, end TOC line
4132 }
4133 }
4137 # count number of headlines for each level
4144 }
4147 }
4148 }
4150 # The safe header is a version of the header text safe to use for links
4152 # Remove link placeholders by the link text.
4153 # <!--LINK number-->
4154 # turns into
4155 # link text with suffix
4156 # Do this before unstrip since link text can contain strip markers
4159 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
4162 # Strip out HTML (first regex removes any tag not allowed)
4163 # Allowed tags are:
4164 # * <sup> and <sub> (bug 8393)
4165 # * <i> (bug 26375)
4166 # * <b> (r105284)
4167 # * <bdi> (bug 72884)
4168 # * <span dir="rtl"> and <span dir="ltr"> (bug 35167)
4169 # * <s> and <strike> (T35715)
4170 # We strip any parameter from accepted tags (second regex), except dir="rtl|ltr" from <span>,
4171 # to allow setting directionality in toc items.
4173 [
4175 '#<(/?(?:span(?: dir="(?:rtl|ltr)")?|sup|sub|bdi|i|b|s|strike))(?: .*?)?>#'
4176 ],
4178 $safeHeadline
4179 );
4181 # Strip '<span></span>', which is the result from the above if
4182 # <span id="foo"></span> is used to produce an additional anchor
4183 # for a section.
4188 # For the anchor, strip out HTML-y stuff period
4192 # Save headline for section edit hint before it's escaped
4196 # For reverse compatibility, provide an id that's
4197 # HTML4-compatible, like we used to.
4198 # It may be worth noting, academically, that it's possible for
4199 # the legacy anchor to conflict with a non-legacy headline
4200 # anchor on the page. In this case likely the "correct" thing
4201 # would be to either drop the legacy anchors or make sure
4202 # they're numbered first. However, this would require people
4203 # to type in section names like "abc_.D7.93.D7.90.D7.A4"
4204 # manually, so let's not bother worrying about it.
4210 # No reason to have both (in fact, we can't)
4212 }
4217 }
4219 # HTML names must be case-insensitively unique (bug 10721).
4220 # This does not apply to Unicode characters per
4221 # http://www.w3.org/TR/html5/infrastructure.html#case-sensitivity-and-string-comparison
4222 # @todo FIXME: We may be changing them depending on the current locale.
4228 }
4230 # Create the anchor for linking from the TOC to the section
4234 // @codingStandardsIgnoreStart
4236 // @codingStandardsIgnoreEnd
4241 }
4243 // @codingStandardsIgnoreStart
4245 // @codingStandardsIgnoreEnd
4250 }
4252 # Don't number the heading if it is the only one (looks silly)
4254 # the two are different if the line contains a link
4258 $numbering
4260 }
4265 }
4267 # Add the section to the section tree
4268 # Find the DOM node for this header
4275 }
4276 }
4280 }
4290 ];
4292 # give headline the correct <h#> tag
4294 // Output edit section links as markers with styles that can be customized by skins
4296 # Put a T flag in the section identifier, to indicate to extractSections()
4297 # that sections inside <includeonly> should be counted.
4305 }
4306 // We use a bit of pesudo-xml for editsection markers. The
4307 // language converter is run later on. Using a UNIQ style marker
4308 // leads to the converter screwing up the tokens when it
4309 // converts stuff. And trying to insert strip tags fails too. At
4310 // this point all real inputted tags have already been escaped,
4311 // so we don't have to worry about a user trying to input one of
4312 // these markers directly. We use a page and section attribute
4313 // to stop the language converter from converting these
4314 // important bits of data, but put the headline hint inside a
4315 // content block because the language converter is supposed to
4316 // be able to convert that piece of data.
4317 // Gets replaced with html in ParserOutput::getText
4324 }
4327 }
4333 }
4337 # Never ever show TOC if no headers
4340 }
4345 }
4350 }
4354 }
4356 # split up and insert constructed headlines
4360 // build an array of document sections
4363 // $head is zero-based, sections aren't.
4368 }
4370 /**
4371 * Send a hook, one per section.
4372 * The idea here is to be able to make section-level DIVs, but to do so in a
4373 * lower-impact, more correct way than r50769
4374 *
4375 * $this : caller
4376 * $section : the section number
4377 * &$sectionContent : ref to the content of the section
4378 * $showEditLinks : boolean describing whether this section has an edit link
4379 */
4383 }
4386 // append the TOC at the beginning
4387 // Top anchor now in skin
4389 }
4397 }
4398 }
4400 /**
4401 * Transform wiki markup when saving a page by doing "\r\n" -> "\n"
4402 * conversion, substituting signatures, {{subst:}} templates, etc.
4403 *
4404 * @param string $text The text to transform
4405 * @param Title $title The Title object for the current article
4406 * @param User $user The User object describing the current user
4407 * @param ParserOptions $options Parsing options
4408 * @param bool $clearState Whether to clear the parser state first
4409 * @return string The altered wiki markup
4410 */
4413 ) {
4416 }
4420 // We still normalize line endings for backwards-compatibility
4421 // with other code that just calls PST, but this should already
4422 // be handled in TextContent subclasses
4427 }
4433 }
4435 /**
4436 * Pre-save transform helper function
4437 *
4438 * @param string $text
4439 * @param User $user
4440 *
4441 * @return string
4442 */
4446 # Note: This is the timestamp saved as hardcoded wikitext to
4447 # the database, we use $wgContLang here in order to give
4448 # everyone the same signature and use the default one rather
4449 # than the one selected in each user's preferences.
4450 # (see also bug 12815)
4458 # Variable replacement
4459 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
4462 # This works almost by chance, as the replaceVariables are done before the getUserSig(),
4463 # which may corrupt this parser instance via its wfMessage()->text() call-
4465 # Signatures
4471 ] );
4473 # Context links ("pipe tricks"): [[|name]] and [[name (context)|]]
4477 // [[ns:page (context)|]]
4479 // [[ns:page（context）|]] (double-width brackets, added in r40257)
4481 // [[ns:page (context), context|]] (using either single or double-width comma)
4483 // [[|page]] (reverse pipe trick: add context from page title)
4486 # try $p1 first, to turn "[[A, B (C)|]]" into "[[A, B (C)|A, B]]"
4498 # if there's no context, don't bother duplicating the title
4500 }
4503 }
4505 /**
4506 * Fetch the user's signature text, if any, and normalize to
4507 * validated, ready-to-insert wikitext.
4508 * If you have pre-fetched the nickname or the fancySig option, you can
4509 * specify them here to save a database query.
4510 * Do not reuse this parser instance after calling getUserSig(),
4511 * as it may have changed if it's the $wgParser.
4512 *
4513 * @param User $user
4514 * @param string|bool $nickname Nickname to use or false to use user's default nickname
4515 * @param bool|null $fancySig whether the nicknname is the complete signature
4516 * or null to use default value
4517 * @return string
4518 */
4524 # If not given, retrieve from the user object.
4527 }
4531 }
4539 # Sig. might contain markup; validate this
4541 # Validated; clean up (if needed) and return it
4544 # Failed to validate; fall back to the default
4547 }
4548 }
4550 # Make sure nickname doesnt get a sig in a sig
4553 # If we're still here, make it a link to the user page
4560 }
4562 /**
4563 * Check that the user's signature contains no bad XML
4564 *
4565 * @param string $text
4566 * @return string|bool An expanded string, or false if invalid.
4567 */
4570 }
4572 /**
4573 * Clean up signature text
4574 *
4575 * 1) Strip 3, 4 or 5 tildes out of signatures @see cleanSigInSig
4576 * 2) Substitute all transclusions
4577 *
4578 * @param string $text
4579 * @param bool $parsing Whether we're cleaning (preferences save) or parsing
4580 * @return string Signature text
4581 */
4587 }
4589 # Option to disable this feature
4592 }
4594 # @todo FIXME: Regex doesn't respect extension tags or nowiki
4595 # => Move this logic to braceSubstitution()
4597 $substRegex = '/\{\{(?!(?:' . $substWord->getBaseRegex() . '))/x' . $substWord->getRegexCase();
4608 }
4611 }
4613 /**
4614 * Strip 3, 4 or 5 tildes out of signatures.
4615 *
4616 * @param string $text
4617 * @return string Signature text with /~{3,5}/ removed
4618 */
4622 }
4624 /**
4625 * Set up some variables which are usually set up in parse()
4626 * so that an external function can call some class members with confidence
4627 *
4628 * @param Title|null $title
4629 * @param ParserOptions $options
4630 * @param int $outputType
4631 * @param bool $clearState
4632 */
4635 ) {
4637 }
4639 /**
4640 * @param Title|null $title
4641 * @param ParserOptions $options
4642 * @param int $outputType
4643 * @param bool $clearState
4644 */
4647 ) {
4653 }
4654 }
4656 /**
4657 * Wrapper for preprocess()
4658 *
4659 * @param string $text The text to preprocess
4660 * @param ParserOptions $options Options
4661 * @param Title|null $title Title object or null to use $wgTitle
4662 * @return string
4663 */
4667 # Guard against infinite recursion
4670 }
4676 }
4682 }
4684 /**
4685 * Create an HTML-style tag, e.g. "<yourtag>special text</yourtag>"
4686 * The callback should have the following form:
4687 * function myParserHook( $text, $params, $parser, $frame ) { ... }
4688 *
4689 * Transform and return $text. Use $parser for any required context, e.g. use
4690 * $parser->getTitle() and $parser->getOptions() not $wgTitle or $wgOut->mParserOptions
4691 *
4692 * Hooks may return extended information by returning an array, of which the
4693 * first numbered element (index 0) must be the return string, and all other
4694 * entries are extracted into local variables within an internal function
4695 * in the Parser class.
4696 *
4697 * This interface (introduced r61913) appears to be undocumented, but
4698 * 'markerType' is used by some core tag hooks to override which strip
4699 * array their results are placed in. **Use great caution if attempting
4700 * this interface, as it is not documented and injudicious use could smash
4701 * private variables.**
4702 *
4703 * @param string $tag The tag to use, e.g. 'hook' for "<hook>"
4704 * @param callable $callback The callback function (and object) to use for the tag
4705 * @throws MWException
4706 * @return callable|null The old value of the mTagHooks array associated with the hook
4707 */
4712 }
4717 }
4720 }
4722 /**
4723 * As setHook(), but letting the contents be parsed.
4724 *
4725 * Transparent tag hooks are like regular XML-style tag hooks, except they
4726 * operate late in the transformation sequence, on HTML instead of wikitext.
4727 *
4728 * This is probably obsoleted by things dealing with parser frames?
4729 * The only extension currently using it is geoserver.
4730 *
4731 * @since 1.10
4732 * @todo better document or deprecate this
4733 *
4734 * @param string $tag The tag to use, e.g. 'hook' for "<hook>"
4735 * @param callable $callback The callback function (and object) to use for the tag
4736 * @throws MWException
4737 * @return callable|null The old value of the mTagHooks array associated with the hook
4738 */
4743 }
4744 $oldVal = isset( $this->mTransparentTagHooks[$tag] ) ? $this->mTransparentTagHooks[$tag] : null;
4748 }
4750 /**
4751 * Remove all tag hooks
4752 */
4757 }
4759 /**
4760 * Create a function, e.g. {{sum:1|2|3}}
4761 * The callback function should have the form:
4762 * function myParserFunction( &$parser, $arg1, $arg2, $arg3 ) { ... }
4763 *
4764 * Or with Parser::SFH_OBJECT_ARGS:
4765 * function myParserFunction( $parser, $frame, $args ) { ... }
4766 *
4767 * The callback may either return the text result of the function, or an array with the text
4768 * in element 0, and a number of flags in the other elements. The names of the flags are
4769 * specified in the keys. Valid flags are:
4770 * found The text returned is valid, stop processing the template. This
4771 * is on by default.
4772 * nowiki Wiki markup in the return value should be escaped
4773 * isHTML The returned text is HTML, armour it against wikitext transformation
4774 *
4775 * @param string $id The magic word ID
4776 * @param callable $callback The callback function (and object) to use
4777 * @param int $flags A combination of the following flags:
4778 * Parser::SFH_NO_HASH No leading hash, i.e. {{plural:...}} instead of {{#if:...}}
4779 *
4780 * Parser::SFH_OBJECT_ARGS Pass the template arguments as PPNode objects instead of text.
4781 * This allows for conditional expansion of the parse tree, allowing you to eliminate dead
4782 * branches and thus speed up parsing. It is also possible to analyse the parse tree of
4783 * the arguments, and to control the way they are expanded.
4784 *
4785 * The $frame parameter is a PPFrame. This can be used to produce expanded text from the
4786 * arguments, for instance:
4787 * $text = isset( $args[0] ) ? $frame->expand( $args[0] ) : '';
4788 *
4789 * For technical reasons, $args[0] is pre-expanded and will be a string. This may change in
4790 * future versions. Please call $frame->expand() on it anyway so that your code keeps
4791 * working if/when this is changed.
4792 *
4793 * If you want whitespace to be trimmed from $args, you need to do it yourself, post-
4794 * expansion.
4795 *
4796 * Please read the documentation in includes/parser/Preprocessor.php for more information
4797 * about the methods available in PPFrame and PPNode.
4798 *
4799 * @throws MWException
4800 * @return string|callable The old callback function for this name, if any
4801 */
4808 # Add to function cache
4812 }
4818 # Case
4821 }
4822 # Add leading hash
4825 }
4826 # Remove trailing colon
4829 }
4831 }
4833 }
4835 /**
4836 * Get all registered function hook identifiers
4837 *
4838 * @return array
4839 */
4842 }
4844 /**
4845 * Create a tag function, e.g. "<test>some stuff</test>".
4846 * Unlike tag hooks, tag functions are parsed at preprocessor level.
4847 * Unlike parser functions, their content is not preprocessed.
4848 * @param string $tag
4849 * @param callable $callback
4850 * @param int $flags
4851 * @throws MWException
4852 * @return null
4853 */
4858 }
4865 }
4868 }
4870 /**
4871 * Replace "<!--LINK-->" link placeholders with actual links, in the buffer
4872 * Placeholders created in Linker::link()
4873 *
4874 * @param string $text
4875 * @param int $options
4876 */
4879 }
4881 /**
4882 * Replace "<!--LINK-->" link placeholders with plain text of links
4883 * (not HTML-formatted).
4884 *
4885 * @param string $text
4886 * @return string
4887 */
4890 }
4892 /**
4893 * Renders an image gallery from a text with one line per image.
4894 * text labels may be given by using |-style alternative text. E.g.
4895 * Image:one.jpg|The number "1"
4896 * Image:tree.jpg|A tree
4897 * given as text will return the HTML of a gallery with two images,
4898 * labeled 'The number "1"' and
4899 * 'A tree'.
4900 *
4901 * @param string $text
4902 * @param array $params
4903 * @return string HTML
4904 */
4910 }
4915 // If invalid type set, fallback to default.
4917 }
4930 }
4936 }
4939 }
4942 }
4945 }
4952 # match lines like these:
4953 # Image:someimage.jpg|This is some image
4956 # Skip empty lines
4959 }
4963 }
4966 # Bogus title. Ignore these so we don't bomb out later.
4968 }
4970 # We need to get what handler the file uses, to figure out parameters.
4971 # Note, a hook can overide the file name, and chose an entirely different
4972 # file (which potentially could be of a different type and have different handler).
4977 # Don't register it now, as ImageGallery does that later.
4984 ];
4987 // We don't want people to specify per-image widths.
4988 // Additionally the width parameter would need special casing anyhow.
4990 }
4999 // look for an |alt= definition while trying not to break existing
5000 // captions with multiple pipes (|) in it, until a more sensible grammar
5001 // is defined for images in galleries
5003 // FIXME: Doing recursiveTagParse at this stage, and the trim before
5004 // splitting on '|' is a bit odd, and different from makeImage.
5022 // check to see if link matches an absolute url, if not then it must be a wiki link.
5029 }
5030 }
5033 // Must be a handler specific parameter.
5037 // Guess not, consider it as caption.
5040 }
5041 }
5044 // Last pipe wins.
5046 }
5047 }
5048 // Remove the pipe.
5050 }
5053 }
5057 }
5059 /**
5060 * @param MediaHandler $handler
5061 * @return array
5062 */
5068 }
5070 # Initialise static lists
5077 ];
5085 }
5086 }
5087 }
5089 # Add handler params
5095 }
5096 }
5099 }
5101 }
5103 /**
5104 * Parse image options text and use it to make an image
5105 *
5106 * @param Title $title
5107 * @param string $options
5108 * @param LinkHolderArray|bool $holders
5109 * @return string HTML
5110 */
5112 # Check if the options text is of the form "options|alt text"
5113 # Options are:
5114 # * thumbnail make a thumbnail with enlarge-icon and caption, alignment depends on lang
5115 # * left no resizing, just left align. label is used for alt= only
5116 # * right same, but right aligned
5117 # * none same, but not aligned
5118 # * ___px scale to ___ pixels width, no aligning. e.g. use in taxobox
5119 # * center center the image
5120 # * frame Keep original image size, no magnify-button.
5121 # * framed Same as "frame"
5122 # * frameless like 'thumb' but without a frame. Keeps user preferences for width
5123 # * upright reduce width for upright images, rounded to full __0 px
5124 # * border draw a 1px border around the image
5125 # * alt Text for HTML alt attribute (defaults to empty)
5126 # * class Set a class for img node
5127 # * link Set the target of the image link. Can be external, interwiki, or local
5128 # vertical-align values (no % or length right now):
5129 # * baseline
5130 # * sub
5131 # * super
5132 # * top
5133 # * text-top
5134 # * middle
5135 # * bottom
5136 # * text-bottom
5140 # Give extensions a chance to select the file revision for us
5145 # Fetch and register the file (file title may be different via hooks)
5148 # Get parameter map
5155 }
5157 # Process the input parameters
5169 # Special case; width and height come in one variable together
5177 }
5178 }
5184 }
5185 }
5186 # else no validation -- bug 13436
5189 # Validate handler parameter
5192 # Validate internal parameters
5197 # @todo FIXME: Possibly check validity here for
5198 # manualthumb? downstream behavior seems odd with
5199 # missing manual thumbs.
5217 }
5219 }
5227 }
5228 }
5233 // use first appearing option, discard others.
5238 # Most other things appear to be empty or numeric...
5240 }
5241 }
5245 }
5246 }
5247 }
5250 }
5251 }
5253 # Process alignment parameters
5256 }
5259 }
5263 # Will the image be presented in a frame, with the caption below?
5269 # In the old days, [[Image:Foo|text...]] would set alt text. Later it
5270 # came to also set the caption, ordinary text after the image -- which
5271 # makes no sense, because that just repeats the text multiple times in
5272 # screen readers. It *also* came to set the title attribute.
5273 # Now that we have an alt attribute, we should not set the alt text to
5274 # equal the caption: that's worse than useless, it just repeats the
5275 # text. This is the framed/thumbnail case. If there's no caption, we
5276 # use the unnamed parameter for alt text as well, just for the time be-
5277 # ing, if the unnamed param is set and the alt param is not.
5278 # For the future, we need to figure out if we want to tweak this more,
5279 # e.g., introducing a title= parameter for the title; ignoring the un-
5280 # named parameter entirely for images without a caption; adding an ex-
5281 # plicit caption= parameter and preserving the old magic unnamed para-
5282 # meter for BC; ...
5285 # No caption or alt text, add the filename as the alt text so
5286 # that screen readers at least get some description of the image
5288 }
5289 # Do not set $params['frame']['title'] because tooltips don't make sense
5290 # for framed images
5293 # No alt text, use the "caption" for the alt text
5297 # No caption, fall back to using the filename for the
5298 # alt text
5300 }
5301 }
5302 # Use the "caption" for the tooltip text
5304 }
5308 # Linker does the rest
5313 # Give the handler a chance to modify the parser object
5316 }
5319 }
5321 /**
5322 * @param string $caption
5323 * @param LinkHolderArray|bool $holders
5324 * @return mixed|string
5325 */
5327 # Strip bad stuff out of the title (tooltip). We can't just use
5328 # replaceLinkHoldersText() here, because if this function is called
5329 # from replaceInternalLinks2(), mLinkHolders won't be up-to-date.
5334 }
5336 # make sure there are no placeholders in thumbnail attributes
5337 # that are later expanded to html- so expand them now and
5338 # remove the tags
5343 }
5345 /**
5346 * Set a flag in the output object indicating that the content is dynamic and
5347 * shouldn't be cached.
5348 * @deprecated since 1.28; use getOutput()->updateCacheExpiry()
5349 */
5355 }
5357 }
5359 /**
5360 * Callback from the Sanitizer for expanding items found in HTML attribute
5361 * values, so they can be safely tested and escaped.
5362 *
5363 * @param string $text
5364 * @param bool|PPFrame $frame
5365 * @return string
5366 */
5371 }
5373 /**
5374 * Accessor
5375 *
5376 * @return array
5377 */
5383 );
5384 }
5386 /**
5387 * Replace transparent tags in $text with the values given by the callbacks.
5388 *
5389 * Transparent tag hooks are like regular XML-style tag hooks, except they
5390 * operate late in the transformation sequence, on HTML instead of wikitext.
5391 *
5392 * @param string $text
5393 *
5394 * @return string
5395 */
5409 );
5412 }
5414 }
5416 }
5418 /**
5419 * Break wikitext input into sections, and either pull or replace
5420 * some particular section's text.
5421 *
5422 * External callers should use the getSection and replaceSection methods.
5423 *
5424 * @param string $text Page wikitext
5425 * @param string|number $sectionId A section identifier string of the form:
5426 * "<flag1> - <flag2> - ... - <section number>"
5427 *
5428 * Currently the only recognised flag is "T", which means the target section number
5429 * was derived during a template inclusion parse, in other words this is a template
5430 * section edit link. If no flags are given, it was an ordinary section edit link.
5431 * This flag is required to avoid a section numbering mismatch when a section is
5432 * enclosed by "<includeonly>" (bug 6563).
5433 *
5434 * The section number 0 pulls the text before the first heading; other numbers will
5435 * pull the given section along with its lower-level subsections. If the section is
5436 * not found, $mode=get will return $newtext, and $mode=replace will return $text.
5437 *
5438 * Section 0 is always considered to exist, even if it only contains the empty
5439 * string. If $text is the empty string and section 0 is replaced, $newText is
5440 * returned.
5441 *
5442 * @param string $mode One of "get" or "replace"
5443 * @param string $newText Replacement text for section data.
5444 * @return string For "get", the extracted section text.
5445 * for "replace", the whole page with the section replaced.
5446 */
5455 # Process section extraction flags
5462 }
5463 }
5465 # Check for empty input
5467 # Only sections 0 and T-0 exist in an empty document
5473 }
5479 }
5480 }
5481 }
5483 # Preprocess the text
5486 # <h> nodes indicate section breaks
5487 # They can only occur at the top level, so we can find them by iterating the root's children
5490 # Find the target section
5492 # Section zero doesn't nest, level=big
5501 }
5502 }
5505 }
5507 }
5508 }
5511 # Not found
5516 }
5517 }
5519 # Find the end of the section, including nested sections
5526 }
5527 }
5530 }
5534 # Write out the remainder (in replace mode only)
5536 # Output the replacement text
5537 # Add two newlines on -- trailing whitespace in $newText is conventionally
5538 # stripped by the editor, so we need both newlines to restore the paragraph gap
5539 # Only add trailing whitespace if there is newText
5542 }
5547 }
5548 }
5551 # Re-insert stripped tags
5553 }
5556 }
5558 /**
5559 * This function returns the text of a section, specified by a number ($section).
5560 * A section is text under a heading like == Heading == or \<h1\>Heading\</h1\>, or
5561 * the first section before any such heading (section 0).
5562 *
5563 * If a section contains subsections, these are also returned.
5564 *
5565 * @param string $text Text to look in
5566 * @param string|number $sectionId Section identifier as a number or string
5567 * (e.g. 0, 1 or 'T-1').
5568 * @param string $defaultText Default to return if section is not found
5569 *
5570 * @return string Text of the requested section
5571 */
5574 }
5576 /**
5577 * This function returns $oldtext after the content of the section
5578 * specified by $section has been replaced with $text. If the target
5579 * section does not exist, $oldtext is returned unchanged.
5580 *
5581 * @param string $oldText Former text of the article
5582 * @param string|number $sectionId Section identifier as a number or string
5583 * (e.g. 0, 1 or 'T-1').
5584 * @param string $newText Replacing text
5585 *
5586 * @return string Modified text
5587 */
5590 }
5592 /**
5593 * Get the ID of the revision we are parsing
5594 *
5595 * @return int|null
5596 */
5599 }
5601 /**
5602 * Get the revision object for $this->mRevisionId
5603 *
5604 * @return Revision|null Either a Revision object or null
5605 * @since 1.23 (public since 1.23)
5606 */
5610 }
5613 }
5617 );
5619 # If the parse is for a new revision, then the callback should have
5620 # already been set to force the object and should match mRevisionId.
5621 # If not, try to fetch by mRevisionId for sanity.
5624 }
5629 }
5631 /**
5632 * Get the timestamp associated with the current revision, adjusted for
5633 * the default server-local timestamp
5634 * @return string
5635 */
5643 # The cryptic '' timezone parameter tells to use the site-default
5644 # timezone offset instead of the user settings.
5645 # Since this value will be saved into the parser cache, served
5646 # to other users, and potentially even used inside links and such,
5647 # it needs to be consistent for all visitors.
5650 }
5652 }
5654 /**
5655 * Get the name of the user that edited the last revision
5656 *
5657 * @return string User name
5658 */
5663 # if this template is subst: the revision id will be blank,
5664 # so just use the current user's name
5669 }
5670 }
5672 }
5674 /**
5675 * Get the size of the revision
5676 *
5677 * @return int|null Revision size
5678 */
5683 # if this variable is subst: the revision id will be blank,
5684 # so just use the parser input size, because the own substituation
5685 # will change the size.
5690 }
5691 }
5693 }
5695 /**
5696 * Mutator for $mDefaultSort
5697 *
5698 * @param string $sort New value
5699 */
5703 }
5705 /**
5706 * Accessor for $mDefaultSort
5707 * Will use the empty string if none is set.
5708 *
5709 * This value is treated as a prefix, so the
5710 * empty string is equivalent to sorting by
5711 * page name.
5712 *
5713 * @return string
5714 */
5720 }
5721 }
5723 /**
5724 * Accessor for $mDefaultSort
5725 * Unlike getDefaultSort(), will return false if none is set
5726 *
5727 * @return string|bool
5728 */
5731 }
5733 /**
5734 * Try to guess the section anchor name based on a wikitext fragment
5735 * presumably extracted from a heading, for example "Header" from
5736 * "== Header ==".
5737 *
5738 * @param string $text
5739 *
5740 * @return string
5741 */
5743 # Strip out wikitext links(they break the anchor)
5747 }
5749 /**
5750 * Same as guessSectionNameFromWikiText(), but produces legacy anchors
5751 * instead. For use in redirects, since IE6 interprets Redirect: headers
5752 * as something other than UTF-8 (apparently?), resulting in breakage.
5753 *
5754 * @param string $text The section name
5755 * @return string An anchor
5756 */
5758 # Strip out wikitext links(they break the anchor)
5762 }
5764 /**
5765 * Strips a text string of wikitext for use in a section anchor
5766 *
5767 * Accepts a text string and then removes all wikitext from the
5768 * string and leaves only the resultant text (i.e. the result of
5769 * [[User:WikiSysop|Sysop]] would be "Sysop" and the result of
5770 * [[User:WikiSysop]] would be "User:WikiSysop") - this is intended
5771 * to create valid section anchors by mimicing the output of the
5772 * parser when headings are parsed.
5773 *
5774 * @param string $text Text string to be stripped of wikitext
5775 * for use in a Section anchor
5776 * @return string Filtered text string
5777 */
5779 # Strip internal link markup
5783 # Strip external link markup
5784 # @todo FIXME: Not tolerant to blank link text
5785 # I.E. [https://www.mediawiki.org] will render as [1] or something depending
5786 # on how many empty links there are on the page - need to figure that out.
5787 $text = preg_replace( '/\[(?i:' . $this->mUrlProtocols . ')([^ ]+?) ([^[]+)\]/', '$2', $text );
5789 # Parse wikitext quotes (italics & bold)
5792 # Strip HTML tags
5795 }
5797 /**
5798 * strip/replaceVariables/unstrip for preprocessor regression testing
5799 *
5800 * @param string $text
5801 * @param Title $title
5802 * @param ParserOptions $options
5803 * @param int $outputType
5804 *
5805 * @return string
5806 */
5809 ) {
5817 }
5819 /**
5820 * @param string $text
5821 * @param Title $title
5822 * @param ParserOptions $options
5823 * @return string
5824 */
5827 }
5829 /**
5830 * @param string $text
5831 * @param Title $title
5832 * @param ParserOptions $options
5833 * @return string
5834 */
5837 }
5839 /**
5840 * Call a callback function on all regions of the given text that are not
5841 * inside strip markers, and replace those regions with the return value
5842 * of the callback. For example, with input:
5843 *
5844 * aaa<MARKER>bbb
5845 *
5846 * This will call the callback function twice, with 'aaa' and 'bbb'. Those
5847 * two strings will be replaced with the value returned by the callback in
5848 * each case.
5849 *
5850 * @param string $s
5851 * @param callable $callback
5852 *
5853 * @return string
5854 */
5873 }
5874 }
5875 }
5877 }
5879 /**
5880 * Remove any strip markers found in the given text.
5881 *
5882 * @param string $text Input string
5883 * @return string
5884 */
5887 }
5889 /**
5890 * Save the parser state required to convert the given half-parsed text to
5891 * HTML. "Half-parsed" in this context means the output of
5892 * recursiveTagParse() or internalParse(). This output has strip markers
5893 * from replaceVariables (extensionSubstitution() etc.), and link
5894 * placeholders from replaceLinkHolders().
5895 *
5896 * Returns an array which can be serialized and stored persistently. This
5897 * array can later be loaded into another parser instance with
5898 * unserializeHalfParsedText(). The text can then be safely incorporated into
5899 * the return value of a parser hook.
5900 *
5901 * @param string $text
5902 *
5903 * @return array
5904 */
5911 ];
5913 }
5915 /**
5916 * Load the parser state given in the $data array, which is assumed to
5917 * have been generated by serializeHalfParsedText(). The text contents is
5918 * extracted from the array, and its markers are transformed into markers
5919 * appropriate for the current Parser instance. This transformed text is
5920 * returned, and can be safely included in the return value of a parser
5921 * hook.
5922 *
5923 * If the $data array has been stored persistently, the caller should first
5924 * check whether it is still valid, by calling isValidHalfParsedText().
5925 *
5926 * @param array $data Serialized data
5927 * @throws MWException
5928 * @return string
5929 */
5933 }
5935 # First, extract the strip state.
5939 # Now renumber links
5942 # Should be good to go.
5944 }
5946 /**
5947 * Returns true if the given array, presumed to be generated by
5948 * serializeHalfParsedText(), is compatible with the current version of the
5949 * parser.
5950 *
5951 * @param array $data
5952 *
5953 * @return bool
5954 */
5957 }
5959 /**
5960 * Parsed a width param of imagelink like 300px or 200x300px
5961 *
5962 * @param string $value
5963 *
5964 * @return array
5965 * @since 1.20
5966 */
5971 }
5973 # (bug 13500) In both cases (width/height and width only),
5974 # permit trailing "px" for backward compatibility.
5983 }
5985 }
5987 /**
5988 * Lock the current instance of the parser.
5989 *
5990 * This is meant to stop someone from calling the parser
5991 * recursively and messing up all the strip state.
5992 *
5993 * @throws MWException If parser is in a parse
5994 * @return ScopedCallback The lock will be released once the return value goes out of scope.
5995 */
6000 }
6005 } );
6008 }
6010 /**
6011 * Strip outer <p></p> tag from the HTML source of a single paragraph.
6012 *
6013 * Returns original HTML if the <p/> tag has any attributes, if there's no wrapping <p/> tag,
6014 * or if there is more than one <p/> tag in the input HTML.
6015 *
6016 * @param string $html
6017 * @return string
6018 * @since 1.24
6019 */
6025 }
6026 }
6029 }
6031 /**
6032 * Return this parser if it is not doing anything, otherwise
6033 * get a fresh parser. You can use this method by doing
6034 * $myParser = $wgParser->getFreshParser(), or more simply
6035 * $wgParser->getFreshParser()->parse( ... );
6036 * if you're unsure if $wgParser is safe to use.
6037 *
6038 * @since 1.24
6039 * @return Parser A parser object that is not parsing anything
6040 */
6047 }
6048 }
6050 /**
6051 * Set's up the PHP implementation of OOUI for use in this request
6052 * and instructs OutputPage to enable OOUI for itself.
6053 *
6054 * @since 1.26
6055 */
6059 }
6060 }